Pursuant to a judicial order, all retarded patients at a state hospital currently treated with major tranquilizers will receive periodic drug holidays with detailed observations of behavior change on medication vs. placebo. Behavior change will be quantified on the basis of frequency counts of objectionable and desired behavior in naturalistic hospital environment. Idiographic analysis will compare behavior under drug and placebo conditions to determine whether drug effects in individual patients is greater than might be expected on the basis of chance variations. Patients will be randomly assigned to sequence of drug and placebo observations periods. The relative "desirability" of behavior under the two conditions will be analyzed by Analyses of Variance and other statistical procedures to determine relative efficacy of each major tranquilizer as compared with placebo.